translation model 0.00296014
word substitution 0.002674693
source word 0.002628937
language model 0.002523657
word accuracy 0.00234947
cipher model 0.002318063
model method 0.00226166
lexicon model 0.002190199
word type 0.00217041
word classes 0.0021543359999999998
model state 0.002112142
sequence model 0.002106182
model lookup 0.00199274
membership model 0.001984945
markov model 0.001963704
model states 0.001962024
underlying model 0.0019410949999999999
translation models 0.0018807910000000001
decipherment problem 0.0017037609999999998
translation task 0.001698118
training problem 0.001696328
machine translation 0.001688007
model 0.00168458
based translation 0.001673836
translation quality 0.001644648
training method 0.001545941
chine translation 0.001535462
training time 0.001487404
hash sampling 0.001460584
decipherment accuracy 0.001449784
language models 0.001444308
substitution cipher 0.0014321960000000002
ciphers decipherment 0.001385755
vocabulary decipherment 0.001383538
based training 0.001367137
final decipherment 0.001352975
partial decipherment 0.001319685
decipherment setting 0.001318514
training procedure 0.001305937
order language 0.001291346
sampling technique 0.001286172
same problem 0.001275855
translation 0.00127556
tagging problem 0.0012471639999999998
simple substitution 0.001224227
substitution ciphers 0.001208174
additional source 0.001185687
cipher tokens 0.001168697
search method 0.001168242
substitution case 0.001158222
new method 0.001141262
first half 0.001127969
language mod 0.001119011
cipher text 0.001118226
ple substitution 0.001117427
probabilistic substitution 0.001115068
language vocab 0.0011045130000000001
first thing 0.0011032820000000001
cessor words 0.001090417
english side 0.001082803
successor words 0.001080414
full search 0.00107775
search approach 0.001077591
multiple cipher 0.001077345
words voc 0.001071089
sand words 0.001069609
context information 0.00106713
this problem 0.00106573
phonic substitution 0.0010592190000000001
many hypotheses 0.00105689
english verbmobil 0.001056135
verbmobil english 0.001056135
cipher task 0.001056041
tic substitution 0.001055799
homophonic substitution 0.001055771
bilistic substitution 0.001055136
cipher sequence 0.0010550849999999999
original corpus 0.001034767
opus corpus 0.001034543
search space 0.0010255260000000001
whole corpus 0.001022467
ing models 0.001020976
text tokens 0.001019957
plaintext tokens 0.001019892
ment problem 0.001014274
second method 0.001013633
verbmobil corpus 0.001005703
sampling 0.00100548
subtitle corpus 0.0010006709999999999
method bleu 9.96062E-4
additional context 9.95328E-4
practical problem 9.8839E-4
ficult problem 9.85115E-4
new parameter 9.83942E-4
beam search 9.80416E-4
decipherment 9.76294E-4
different cardinalities 9.750679999999999E-4
different approximations 9.750679999999999E-4
training 9.68861E-4
many successor 9.58617E-4
